I'm not sure what the notation you're using represents, but it looks like you're talking about matrices of a known size. What that's an example of is dependent types, where types can be parameterized by values, and not just types. For example, a (2D) matrix is parameterized by what it contains, and its dimensions. The former is a type but the latter are values. Without dependent types (or some subset of their capability) you can't represent this matrix type; you'd just have to check at runtime that no out-of-bounds access was occurring, so no; I don't think you could put this in a conventional language (without some crazy gymnastics, perhaps involving macros or similar). However, with dependent types, this is absolutely possible to statically guarantee that no out-of-bounds access will occur at runtime.